A Unified Communication Platform for Enriching andEnhancing Presentations with Active Learning Components

Reinout Roels, Christophe Vermeylen and Beat SignerWeb & Information Systems Engineering Lab

Vrije Universiteit BrusselPleinlaan 2, 1050 Brussels, Belgium{rroels,cvermeyl,bsigner}@vub.ac.be

Abstract—There exist various audience response sys-tems (ARS) and other standalone solutions to enrich presenta-tions with supplementary active learning content. We propose aunified communication platform for enriching and enhancingpresentations with active learning components where third-party solutions can easily be integrated via data adapters,activity plug-ins as well as a multi-directional communicationchannel based on a publish-subscribe pattern. We present howthe MindXpres presentation platform has been extended with acommunication module and discuss a number of activity plug-ins that have been realised. The presented unified communica-tion platform in combination with the MindXpres presentationtool represents an ideal platform for the investigation andrapid prototyping of new forms of presentations that supportenhanced forms of interaction between a presenter and theiraudience.

Keywords-Presentation; audience response system; activelearning

I. INTRODUCTION

Over the last two decades we have seen the rise of activelearning as an alternative to the classic lecture setting whereoften little interaction happens between the instructor andtheir audience. As defined by Bronwell and Eison [1], activelearning is about “involving students in doing things andthinking about the things they are doing”. As technology hasbeen evolving, different forms of student engagement haveadapted alongside and are increasingly offered in the formof so-called digital audience response systems (ARS). How-ever, available hardware, applications and forms of interac-tion are quite diverse and frequently offered as standalonebundles. One issue that we address with the presented unifiedcommunication platform is that the functionality of currentresponse systems is often too dependent on specific hardwaremanufacturers. Existing hardware is frequently bundled withsoftware providing some fixed functionality. Unfortunately,a user is limited to the offered functionality and is typicallynot able to use the hardware in other settings. For similarreasons, it is often not possible to utilise hardware fromdifferent manufacturers with the same audience responsesystem for uniform data collection. In order to addressthese challenges, we present a unified approach addressingthe technical challenges involved in the coordination of

hardware, media and the involved parties during lectures andpresentations. We start by discussing some related work. Wethen show how the MindXpres [2], [3] presentation tool hasbeen extended based on a number of identified requirements,in order to address the shortcomings of existing solutions.The presented architecture of our unified communicationplatform allows for arbitrary combinations of hardware to beused together with some plug-in-based audience interaction.The flexible integration of new devices together with thehighly extensible MindXpres platform makes the presentedsolution a valuable asset as a research platform for pro-totyping novel audience-driven functionality or interactionmodalities for future presentation solutions.

II. BACKGROUND

Since the introduction of the term active learning byBronwell and Eison [1], this category of instructional ap-proaches has been thoroughly evaluated [4]–[7]. In general,the literature shows that a higher level of engagementincreases attendance as well as motivation and aids inthe learning process [8]. For example, a student who isforced to participate in small quizzes during the lecturewill remember more content. Active learning is of coursenot limited to quizzes only. Students can, for instance, alsobe further engaged via discussions, exercises, brainstorming,simulations or games.

One of the more common forms to involve the audience ina presentation is through the use of so-called clicker systems.These are small portable handheld devices with a limitednumber of buttons, allowing its holder to select and submita choice during polls or quizzes. Turning Technologies1

and i>clicker2 are only two of the various manufacturersof commercial clicker solutions. Due to the diversity intechnical characteristics (e.g. the used wireless technologyor communication protocols) the devices share a commonshortcoming in the sense that they can only be used withthe software supplied by the vendor and cannot be mixed.This also implies that the usefulness of the hardware is

1http://www.turningtechnologies.com2http://www.iclicker.com

determined and sometimes restricted by the activities andfeatures offered by the corresponding vendor’s software.

With technology becoming cheaper and more ubiquitous,we also see an increasing use of mobile devices, includingsmartphones, tablets or laptops, for audience-driven inter-action. Standard Internet connectivity is commonly usedto deliver activities and send back results which reduceshardware requirements, meaning that a touch screen-enableddevice with a modern web browser is often sufficient. In thiscategory of audience response-driven solutions, various com-mercial products such as Infuse Learning3 or TodaysMeet4

are available. Nevertheless, also for these solutions we canidentify a number of issues. First of all, functionality isfragmented across solutions. Some systems focus on deliv-ering activities and gathering results while others providecommunication channels for discussions and brainstormingsessions. Furthermore, while the use of mobile devices iscertainly beneficial, the reviewed systems do not offer thefunctionality to integrate third-party hardware such as digitalpens5 or the previously mentioned clicker systems. Finally,while clicker systems sometimes offer plug-ins for popularpresentation tools, these web-based systems come in theform of a web application and require the instructor toconstantly switch between the audience response systemand a presentation tool such as PowerPoint. Note that weidentified a number of other systems using different typesof hardware, such as the TI-Navigator6 or Bic Education7,showing similar limitations.

III. REQUIREMENTS

The shortcomings that we identified in the previous sec-tion can be traced back to three aspects of classroom systemsthat have received little attention in the past. These threeaspects include hardware interoperability, extensibility ofthe offered functionality and the integration of rich mediaactivities that go beyond simple poll results in presentationtools. We proceed by outlining some requirements to ad-dress these shortcomings which also form the basis for theimplementation of our unified communication platform.

A. Multi-directional Communication Flows

Initially, audience response systems were seen as a com-munication channel that allows students to send informa-tion, such as the answers selected during a quiz, to theinstructor. Later, the importance of communication betweenstudents has also been recognised and adopted in classroomcommunication systems [9], [10]. In order not to limitthe potential of an extensible communication platform, wedeem it necessary to support bidirectional communication

3http://www.infuselearning.com4https://todaysmeet.com5http://www.anoto.com6http://education.ti.com7http://www.bic-education.com

between any of the involved parties. Additionally, it shouldalso be possible to create partitions of parties (e.g. groupsof students) via the underlying communication framework.Different settings require different roles such as anonymoususers, identified users, instructors or content moderators. Wetherefore propose the adoption of a customisable role modelfor the distribution of authority and the corresponding linkedfunctionality during activities.

B. Hardware Interoperability

As mentioned previously, there are many different formsof input devices for electronic response systems, includ-ing clicker systems, digital pens [11] and proprietary orgeneral purpose mobile devices such as tablets, laptops orsmartphones. Often these systems come with proprietarysoftware that dictates the supported features. A contributionof our proposed platform is the elimination of this type ofvendor lock-in. We discuss an extensible mechanism thatallows different input and output devices to be integrated inthe system with several important implications. First of all,institutions that have already invested in particular hardwarewill be able to use it with the proposed platform and canbenefit from new functionality not offered by the originalsoftware. Second, since regular mobile devices can be used,we foresee to reduce the costs of full-fledged audienceresponse systems. Last but not least, a common interfacealso allows arbitrary combinations of device types to beused together whenever an activity supports this form ofinteraction.

C. Activity Plug-ins

When using third-party hardware, one is often limitedto use the software suite that comes with the product. Wepropose the modularisation of functionality and scenariosinto so-called activity plug-ins in order to allow third partiesto contribute. To further simplify the development process,the communication framework has to provide the necessaryabstractions via an application programming interface (API).Thereby plug-ins can make use of connected hardware andthe various communication flows presented earlier.

D. Integration with Presentation Tools

In order to provide a seamless learning experience, theintegration of an audience response system within existingpresentation tools can only be beneficial. Additionally, werecognise that a presentation’s lifetime does not necessarilyend after a lecture. Even though it was not the originalpurpose of slide decks, they sometimes form a core part ofa student’s available study material. As educational materialis evolving, the post presentation phase might become moreimportant and requires extra attention. Content is becomingincreasingly interactive and non-linear, and a lot of informa-tion is lost when student’s are offered static slide decks asstudy material. As a final guideline, we therefore state that

data gathered during a presentation should be replayable ata later stage. For example, the study material received bythe student could contain the path that was traversed bythe instructor, their annotations (as well as those by otherstudents) or the questions and answers that were generatedat a given point in time.

IV. UNIFIED COMMUNICATION PLATFORM

Based on the principles presented in the previous sec-tion, we have developed a functional prototype of theproposed communication platform by extending the existingMindXpres presentation tool [2], [3]. One of the character-istics of MindXpres is the extensible data-driven approachfor visualising content. Presentations are defined in a LATEX-like language and similar to LATEX, the layout of contentis handled automatically based on themes and plug-insafter compilation to a portable HTML5-based presentationbundle. Plug-ins are implemented in JavaScript and mayadd content or attach event handlers to the DOM treein order to visualise the data that they are responsiblefor. Unlike existing declarative languages for presentations,MindXpres is geared towards modern presentations wherenavigation becomes increasingly non-linear and content isricher and more interactive. As most of the visualisation andnavigation is realised via plug-ins, the MindXpres programcode consists of little hardcoded functionality. Rather, itoffers powerful abstractions such as a graphics engine,where plug-ins can be used to integrate novel presentationmedia types with as little effort or restrictions as possible.Similar to the existing graphical abstractions, we have nowextended the core of MindXpres with a communicationmodule. The goal is to provide the necessary abstractionsthat MindXpres plug-ins do not need to worry about thetechnical challenges related to the connectivity in audienceresponse systems. The provided interface allows plug-ins tointeract with audience hardware in a uniform way and makeuse of the multi-directional communication flows describedearlier. Therefore, the plug-ins can focus on implementingspecific audience-driven activities.

A. Network Infrastructure and Central Access Point

On the network infrastructure level, our solution is basedon a small but powerful computer that acts as a centralhub for audience devices and instances of a MindXprespresentation. For our prototype we used an Intel Next Unitof Computing Kit (NUC) equipped with high end WiFi andBluetooth modules for wireless connectivity. Although wemainly focussed on wireless devices, the access point alsoprovides USB, HDMI and Ethernet interfaces which areequally valid means of adding new devices to the platform.Our setup consists of off-the-shelf components but is by nomeans defining the platform. Since the implementation ofpreviously presented principles mainly contributes on thesoftware level, one could easily switch the access point

for a similar device with other hardware or adapt it fordifferent scenarios. Mobile devices, which can act as input aswell as output device, simply connect to the WiFi networkthat is broadcasted by the access point. After connectingto the access point, a user is presented with a MindXprespresentation instance and forms part of the network ofdevices. In addition to WiFi, Bluetooth can be used as away to add devices such as digital pens or headsets to thenetwork.

B. Unified Hardware Interface

In order to provide a solution for hardware interoper-ability, we need to support a variety of communicationprotocols and channels. As a first part of the communi-cation module, we introduce the concept of data adapterswhich translate device-specific communication formats toa uniform representation before the core and its plug-inscan handle the data as illustrated in Figure 1. If an inputdevice already conforms to the proposed interface (e.g. aweb application developed for our platform) it can directlyuse the interface to pass input data to the communicationmodule via the WebSocket protocol. For third-party devicesthat do not directly interface with the MindXpres core, adata adapter that performs the translation between a device-specific protocol and the access point’s WebSocket-basedprotocol has to be developed. The access point can therebyact as a container for JavaScript-based adapters but they canalso be implemented as separate processes and in differentprogramming languages on the access point or even onother machines. For instance, a dedicated computer couldbe used to perform processor-intensive gesture recognitionon a video feed and pass the detected higher-level gesturesto the central hub via the WebSocket interface. Data adaptersare often bi-directional but in some cases, for instance foroutput-only devices such as large LCD displays or projectorsfor the delivery of assessments, they might also be limitedto one-way translation.

Figure 1: Access point as a bridge between hardware

C. Inter-Presentation Information Flow

While the access point acts as a bridge between differenthardware components, we also have to provide a mech-anism to orchestrate the flow of data between connecteddevices in order to provide more than simple broadcast-based communication. We decided to base our solution onthe publish-subscribe software pattern and offer a publish-subscribe mechanism on the access point. Devices that arerunning an instance of the presentation and its associatedplug-ins can make use of a number of abstractions to sendevents to specific parties (publish) or they can ask the accesspoint to receive specific events generated by plug-ins runningon other devices (subscribe). Our abstractions further enablethe targeting of specific roles and partitions as suggestedearlier when discussing the requirements. Any unidentifieddevice is given the role of an anonymous student by defaultbut a user can then identify themselves to obtain a role withmore authority. All users are served the same instance of theMindXpres presentation with the same set of plug-ins but thecurrently acquired role can be used by plug-ins to changetheir behaviour and the offered functionality accordingly. Aninstructor might, for example, be able or disable a quiz whilethis feature will not be offered to a student.

D. Unified Communication Platform in Use

As an illustration of how the presented components worktogether, we would like to provide a detailed description of afeature that we call navigation mirroring and which will alsobe used by the use cases in the next section. The generalidea is that a presenter’s navigational state represented bythe currently projected content is also shown on the screensof devices forming part of the network. If the presenternavigates to another piece of content, the content should alsobe updated on other devices forming part of the network.We have realised this functionality via a single MindXpresplug-in.

Figure 2: View mirroring via publish-subscribe mechanism

In Figure 2 we illustrate how the plug-in makes use of thepublish-subscribe system to synchronise the views. Step (1)shows how devices that are interested in receiving a mirroredview (i.e. student A and student B) may subscribe to eventsrelated to navigation. When executing with the anonymousor student role, the plug-in simply acts as a listener. Step (2)highlights how the mirroring plug-in executed under the

instructor role publishes navigational changes made by theinstructor. The provided abstraction automatically deliversthe event that is generated on the instructor’s device to theaccess point which then forwards it to any other devices thathave subscribed for this type of event. Once the subscribingplug-ins receive an event, they only need to update their viewbased on the information attached to the event. Note that thisbasic idea has been extended to allow audience members totake over navigation if they are given the permission. Instep (3) we illustrate how a mirroring plug-in running underthe student role (i.e. student C) can also trigger view updateson subscribed devices if given permission.

V. USE CASES

After having discussed the technical details, we wouldlike to illustrate the potential of our unified communicationplatform by describing some of the realised activity plug-ins.

Polls and Quizzes: The poll or quiz plug-in allows audi-ence members to use their hardware to select one of multiplepossible answers for a given question. The results caneither be collected anonymously or traced back to specificusers. The plug-in also provides the instructor with a real-time overview of the incoming answers as highlighted inFigure 3 (c). Questions and answers can be hidden by eitherblurring them out or by keeping them completely invisibleuntil the instructor opens the quiz for input.

Navigation Mirroring: The mirroring plug-in allows otherdevices to see a copy of the currently presented content ontheir own devices as shown in Figure 3 (b). If the presenternavigates to different content, this is immediately propagatedto all connected devices. Note that audience members canenable or disable this functionality at any time.

Navigation Takeover: As it is often inappropriate to inter-rupt a presenter during a presentation to ask a question, mostpeople keep their questions until the end of a presentation.However, since the relevant content is likely not displayedanymore, the presenter might be asked to navigate backto the corresponding slide. This can lead to problems ifthe audience member does not know the slide number orthe presenter does not immediately understand what theaudience member wants to see. To deal with this issue, wedeveloped a plug-in that lets audience members control thepresentation. In order to avoid any abuse, audience membersneed to digitally request permission or the presenter has toselect them manually.

Content Discussions: In order to allow audience membersto debate about the content during or after the presentation,we have developed a plug-in that allows them to providecomments or ask questions about individual slides. In thespirit of social media, users can mark a question or commentas interesting which ranks this question higher than otherquestions. The presenter is offered an overview of theaudience feedback as illustrated in Figure 3 (a) which allowsthem to identify the most popular unanswered questions or

(a) Questions (b) Navigation mirroring (c) Polls and quizzes

Figure 3: Some of the activities implemented as MindXpres plug-in

to spot parts of the presentation with a lot of activity. Theplug-in can not only be used to collect questions during thepresentation for answering them afterwards, but it can alsoserve as a discussion platform in the post-presentation phase.

Annotations: The ability to use different types of hardwareas input media has interesting consequences for annotatingthe content. Annotations can be made with the mouse,keyboard, touch-enabled devices or digital pens. Annotationsby the presenter are relayed to other instances of the pre-sentation in real time or they can be kept hidden if desired.Future work includes the ability to allow audience membersto share annotations with other audience members.

VI. DISCUSSION AND FUTURE WORK

There is no shortage of audience response or classroomconnectivity systems and we do not claim that the presenteduse cases are unique to the MindXpres presentation platform.However, our contribution is a unifying communicationplatform making this category of systems more open andextensible on the software as well as on the hardware level.The presented architecture allows a broad range of hardwareto be unified for usage within the extensible MindXpres pre-sentation solution. Immediate benefits include the possibilityto give existing proprietary hardware a new purpose duringlectures and classes and use them for a wide selection ofpluggable audience-driven activities. However, in the longterm, MindXpres shows potential as a rapid prototypingplatform for investigating innovative presentation features.

The contribution of this paper is mainly on the architec-tural level but future work includes the evaluation of theimplemented use cases in terms of their usability. Also onthe technical side there is still room for improvement. Thesingle access point results in a star-shaped network topologywith a single point of failure and limited scalability. We aretherefore currently investigating WebRTC-based P2P con-nectivity as an alternative. Furthermore, we are looking intosupporting virtual classrooms which involves the extensionof our system’s reach beyond the local network in order toconnect remote users and devices. Given that we achieved auniform representation of the data flow between users and

devices, we also see potential for integrating functionalityrelated to learning analytics. A core MindXpres module forgathering and processing analytics would provide immediateaccess to all the standardised information shared betweenusers and devices, resulting in a more detailed analysis.

REFERENCES

[1] C. C. Bonwell and J. A. Eison, Active Learning: CreatingExcitement in the Classroom. Jossey-Bass, February 1991.

[2] R. Roels, “MindXpres: An Extensible Content-driven Cross-Media Presentation Tool,” Master’s thesis, Vrije UniversiteitBrussel, 2012.

[3] R. Roels and B. Signer, “An Extensible Presentation Tool forFlexible Human-Information Interaction,” in Demo Proceed-ings of HCI 2013, 27th BCS Conference on Human ComputerInteraction, London, UK, September 2013.

[4] R. Hake, “Interactive-Engagement Versus Traditional Meth-ods: A Six-Thousand-Student Survey of Mechanics TestData for Introductory Physics Courses,” American journal ofPhysics, vol. 66, no. 1, January 1998.

[5] C. Hoellwarth and M. J. Moelter, “The Implications of aRobust Curriculum in Introductory Mechanics,” AmericanJournal of Physics, vol. 79, no. 5, 2011.

[6] J. P. Holdren et al., “Engage to Excel: Producing One MillionAdditional College Graduates With Degrees in Science, Tech-nology, Engineering, and Mathematics,” Executive Office ofthe President of the United States, February 2012.

[7] M. Prince, “Does Active Learning Work? A Review of theResearch,” Journal of Engineering Education, vol. 93, no. 3,July 2004.

[8] R. H. Kay and A. LeSage, “Examining the Benefits andChallenges of Using Audience Response Systems: A Reviewof the Literature,” Computers & Education, vol. 53, no. 3,November 2009.

[9] E. Judson and D. Sawada, “Learning from Past and Present:Electronic Response Systems in College Lecture Halls,” Jour-nal of Computers in Mathematics and Science Teaching,vol. 21, no. 2, April 2002.

[10] R. J. Dufresne, W. J. Gerace, W. J. Leonard, J. P. Mestre,and L. Wenk, “Classtalk: A Classroom Communication Sys-tem for Active Learning,” Journal of Computing in HigherEducation, vol. 7, no. 2, 1996.

[11] A. Narduzzo and N. R. Parmar, “Chalk, Talk, Digital Pensand Audience Response Systems - Combining Tradition andTechnology to Improve Maths Learning,” Poster, InnovationsDay, Bath, UK, May 2011.